Many defined benefits pension plans are underfunded and the gap is often large compared to the sponsoring company’s net worth. Beneficiaries are exposed to massive losses when pension schemes are ended and schemes with the worst funding records are those most at risk.
Unless a company is bankrupt, it is obliged to meet any shortfall in the pension plan. But when companies go bankrupt, only the assets of the fund are usually available to pay pensions and beneficiaries lose heavily when schemes are underfunded.
Individuals who rely on pensions often face financial catastrophe when their schemes default. The risks, which arise from underfunding and bankruptcy, are ones which they cannot currently reduce by diversification. They bear a large company-specific risk.
Members of schemes are in effect unsecured and off-balance captive lenders to the company. It is almost impossible to imagine a less suitable investment for them than such a risky and undiversified claim. A market in pension claims would enable beneficiaries to reduce their specific risk by diversification and shift the systematic risk to appropriate investors.
The result would be an increase in general welfare. On top of this, trading pension claims would assign a fair value to each claim, establishing a benchmark for the performance of pension funds. A market in claims would also allow the premiums of State sponsored insurance arrangements to be risk based and priced on an arm’s-length basis.

With the benefit of an accommodating legal and tax environment it should be possible to form collateralised pension claim obligations (CPCO) in a similar way to collateralised debt obligations (CDO). We suggest that governments should ensure that such an environment exists.
To value defined pension liabilities, assumptions must be made about longevity, salary growth and the discount rate. Existing regulations and practices allow considerable latitude about these assumptions, and thus on the level of funding.
According to the US Pension Insurance Data Book, only 3.3% of claims involved in plans terminating between 1975 and 2003 had a funded ratio of more than 75%, and more than half of all claims faced a funding-level of less then 50%.
Sponsors decide on the amount of contributions to the pension fund made by the company every year. Casual discussions with practitioners in the field reveal that approximately two thirds of the UK defined benefits pension schemes have the yearly contribution decided by the sponsor, not by the trustees. Furthermore, the pension plan’s asset allocation is often determined by the sponsor.
By some non-innocent twist of reasoning, actuaries and consultants often value the liabilities by discounting them above the risk free rate. The common justification is that the expected return on plan assets is a long-term measure and therefore consistent with the long-term nature of the plan liabilities. This is not, however, a justifiable assumption. There are two reasons for this. One applies generally, the other applies intermittently. The general reason is that while risky assets, such as equities, should give higher returns than bonds, they will also carry higher risks. The correct funding level for a fund should take into account both the risks being run and the expected return. No investment policy can produce higher returns without involving higher risks. Therefore, the contribution rate should not depend on the investment policy.
This point can be readily demonstrated by reductio ad absurdum – ie by taking the opposite view to its logical extreme. If it is assumed that equities will give a higher return than bonds, then investing in leveraged equities will give an even higher return than equities. On this set of assumptions, any level of contributions could be shown to be adequate, provided risk is ignored.
While in general, therefore, the contribution rate should be independent of the investment policy, there is an additional reason for trustees to insist on this today. This is because the return on equities is not stable over time and the fluctuations are broadly predictable.
It has long been established among financial economists that after prolonged periods of above average returns, such as those of the past 10 to 30 years, equity returns will be poor.
Investing in equities is thus a way for the sponsors to reduce the assessed value of the pension liabilities and thus the required rate of contribution, while letting the beneficiaries bear most of the risks. In current circumstances, both management and shareholders are encouraged to put scheme members at risk by investing in equities.

Much of the shortfall in pension assets has occurred because of the way the assets have been managed. Sponsors and trustees need some yardstick against which they can evaluate the investment management. In the absence of a market in pension claims, the applied benchmark has become the performance of funds themselves and is thus an asset and stock market related benchmark, which is independent of the liabilities of the plan.
What is needed, therefore, is a benchmark that captures the value of the liabilities on a daily basis and against which the performance of pension fund managers could be measured. That would provide them with a strong incentive to manage assets in such a way that they would not become under-funded again.
To be meaningful, however, such a benchmark can only be constructed from actual prices of trades. A prerequisite of benchmark construction is the organisation of a market for pension liabilities. A market in pension claims would provide a market-based judgment of performance, which would take into account the liabilities of the scheme. If the sponsors wished to follow a higher risk investment policy, this would not be penalised by the market if the funding of the scheme were sufficient to offset the risks being run.
Markets in pooled liabilities, such as CDOs, have become increasingly popular in recent years. Packaging and making markets in traded pension claims should be possible and would greatly increase welfare.

The assets of the CPCO would be pools of pension claims handed over by beneficiaries.
The participant would get a claim on a CPCO tranche, which he would select according to his risk affinity and in the context of his overall wealth, in return for the personal pension claim he trades in.
At present the owners of the assets (the beneficiaries) have a high specific risk which they are unable to reduce by diversification. Even if total defaults did not fall, the impact would be spread among many more investors. Since the loss of 20% of a pensioner’s income represents more than twice the loss of welfare involved in a 10% loss, diversification would improve welfare. Moreover, insurance against systematic risk could be bought and sold, enabling pensioners to adjust for their individual risk profiles. If other investors, not just pension beneficiaries were allowed to own part of the CPCOs, risk-averse beneficiaries would be able to reduce the systematic risk of more general pension defaults, which could be redistributed optimally.
A market for pension claims would increase the pressure on companies to fund their plans in full. Companies with large pension deficits will thus resist having their scheme members participate in a CPCO.
Governments should be interested in establishing a pension claim market, since benefits to social welfare as well as those to the government itself are considerable. Governmental agencies such as the Pension Benefit Guarantee Corporation (PBGC) are charged with compensating the beneficiaries if a plan ends without sufficient money to pay all benefits, and is not compensated adequately by the insured companies.
Governments have had to provide guarantees that reduce an individual’s investment risk, such as a fixed minimum rate or a minimum rate relative to the performance of other pension funds. In general, government guarantees cause a large burden on the taxpayer, in addition to potential agency problems.
Public pressure and government help would thus contribute to an increase in welfare and a reduction in pension problems.
The form of securitisation that we are proposing will not die with the growth of defined contribution plans. Even then, a household will be better off holding the class of pension CDO securities that it wishes to hold. In this way, it will be able to tailor the risk it bears to its preferences.
Providing insurance to the common man is the raison d’être of financial institutions.
There is no reason for which they should abandon that role when it comes to pensions.
Bernard Dumas, Ian Edwards and Juerg Syz are with INSEAD; Andrew Smithers is with Smithers & Co